The VTC exercise offers an experiential opportunity to behave, cope, deal with the exigencies of news automation (robots, feeds, agents etc). The impact of automated journalism is driving a separation between the values and principles of journalism (ref: http://ethicaljournalismnetwork.org/en/contents/5-principles-of-journalism) and actors desire to control or manipulate information.
In terms of assumptions surrounding the 'impact' of big data we see claims in areas such as automated journalism, that they are having an impact in driving societal responses, not merely at informing the populace.
News feeds aggregating social media activity in or near 'real-time' may act as the first sign or signal of a major event, incidents such as: the Paris attacks of Nov 13 2015 (link to an article on social media unfolding), the Aug 12 2015 Tianjin explosions, the Nepal earthquake of April 25 2015, the Charile Hebdo attacks of Jan 7 2015, or the Mar 11 2011 TÅhoku earthquake and tsunami that led to the Fukushima nuclear disaster.
The following scenario simulates a major event taking place in the Netherlands. Your role, as a member of a team of crisis-mappers, is to report on what is actually happening.
You are tasked with creating a description (map, timeline, narrative) of the unfolding situation in order to help first responders and emergency services, to establish what is happening, when and where it occurred.
Process:
Consume news and information (collapsing on train)
Theory - hypothesis formation (from a person suffering heart attack to gas attack on train)
Trust/test - Narrative coherence (triangulated independent sources)
| Legitimacy Matrix | Verified/True | Uncertain/Doubtful | Unknown/unverified |
|---|---|---|---|
| Immediate result | :-) | :-\ | :-| |
| Delay (mins - hrs) | :-\ | :-| | :-( |
| Delay (days) | :-| | :-( | :-( |
Team A: https://sites.google.com/site/thecrisismappers/
Team B: https://sites.google.com/site/workshoptestice/
Team C: https://sites.google.com/site/theptolemysociety/
VTC Team C
To following is my real-time analysis of key points in the time-line needing verification analysis:My focus is on data trustworthiness, accuracy, relevance. All statements must stand up to source-checking protocol. Focus on actionable, reliable evidence/intelligence.
Comments:
Unstructured data monitoring. Not necessarily any useful hashtags in play at the outset, how to monitor efficiently?
Positive (instances, traces, events) data sources:
Standard search term agents in continuous circulation (e.g. "bang", "gas", "shooting", "collapse". Traffic/chatter volumes.
Image detection agents (e.g. crowds running, signature image vibration in webcams, image saturation)
Audio agents (e.g. sound pulses, signature noises).
Negative (unexpected absence of information) data sources:
Dropped data feeds, blackouts, localised absence of traffic data, failed checkins.
8:00
Mac collapses
Train delays
Train commotion
Train gas leak
Gas smell schiphol
8:02
8.05
Police and medics at rail (pic)
London calling ignore:
DutchTrainCo announces avoid Schiphol.
KevinChan bomb or somthing
Paul smells like down there (actually there?)
8:13
Dutch_News 2 passengers critical arrive hospital.
8:14
4 in hospitcal
Person: Someon ecolplapsedin m ctrain
Amar RT
8:18
Dutch_News: Victims redirected to other hospital (more?)
8:19
#needcoffeenow
8:21
Dutch_police please avoid Schipol
Dutch_police Hague CS to be avoided.
8:30
(fromiPhone?)
Dutch_News prep emergency patient surge
8:32
Dutch_News traffic
8:34
Dutch_Police suspect sarin
8:40
Boyfriend was terrorist
Live stream commentary:
8:10
We Video repeats much of same info from twitter feed (stock footage?)
8:15
Not live video but appears to be clips from today, and some photos
8:40
Sarin announcement, victims and hospital congestion
Appendix:
A report is verified if confirmed by:
- Mainstream news media (i.e. NOS)
- Other reliable news sources (i.e. “Dutch News”)
- Report on same event that has been verified already by two independent sources
- Information from the Dutch Governement or other official organizations or agencies (Police or Fire)
- Twitter communication (proceed with caution!!!) - twitter should only be verified if you have twitter communication with the person who submitted the original tweet and they can provide additional sources of confirmation that meet verification criteria. If you are unsure, ask. Paste any new info into the description box of the report. If the issue is urgent/an emergency, communicate the issue immediately to your team leaders via skype and/or email.
A report is NOT verified if confirmed by:
- Retweets or multiple tweets on the same event
- Multiple unverified reports that say the same thing
- Report is of “allegations” or other unverified information
- There no direct link to the source
Verification Matrix
(Inspired by SBTF Verification Matrix)
This matrix contains different types of information and their verification needs. This is not an exhaustive list; new items will likely be identified to suit the specific emergency situation. However, this list provides a good idea of the types of information that will probably be requested. It will also show which layers can be built prior to an emergency, to speed up the creation of the maps.
Urgency (can depend on the scenario):
1 = Very Urgent
2 = Urgent
3 = Not Urgent
Verification Level (can depend on the scenario):
1 = only information from Trusted Sources should be considered verified (e.g. government websites, major organizations, etc.)
2 = if 3 separate sources report the same thing, the information can be considered verified.
3 = no verification required
Info Type
|
Layer Details
|
Urgency Level
|
Verification Level
|
I. Open Source Items
| |||
Locality Information
|
Shelters, food distribution centres and relief stockpiles, as identified by local authorities
|
1
|
2
|
Local population density
|
3
|
2
| |
Hotels, Supermarkets, Groceries, Pharmacies
|
1
|
2
| |
Local First Response Centres: Hospitals, Fire Stations, Police Stations, etc.
|
2
|
1
| |
Key Local Government Buildings
|
2
|
1
| |
‘Gathering’ places.
|
2
|
2
| |
Transportation Lines and Points: Roads, Rails, Airports, Ports, bus terminals, etc. Any closures, road blocks, check points, etc…
|
1
|
2
| |
Events, Incidents, Hazards
|
Locations of incidents - Protests, explosions, assassinations, blockades, etc.
Extent of natural disaster
· Flood extent, earthquake damages, etc.
· Public health threats – e.g. Dengue Fever Zone
· Trajectory
|
1
|
2
|
Hazards & Risk (e.g. Fault Lines, Low-lying flood plains, microzonation maps, storm trajectories, etc.)
|
1
|
2
| |
Bridges or tunnels at risk of structural instability or collapse.
|
1
|
2
| |
Reported deaths/injuries
|
1
|
1
| |
II. Closed Sourced Items
| |||
Alternate routes to main transportation hubs
|
1
|
2
| |
Areas to avoid
|
1
|
2
| |
Planned Evacuation Routes/Points
|
1
|
2
| |
Planned Assembly Points
|
1
|
2
| |
Forecasts & Risk Assessments
|
1
|
2
| |
Identified ‘hazardous areas’
|
1
|
2
|
